CN113239469A - Structure optimization method, device, equipment and storage medium for vehicle body parts - Google Patents

Structure optimization method, device, equipment and storage medium for vehicle body parts Download PDF

Info

Publication number
CN113239469A
CN113239469A CN202110662295.9A CN202110662295A CN113239469A CN 113239469 A CN113239469 A CN 113239469A CN 202110662295 A CN202110662295 A CN 202110662295A CN 113239469 A CN113239469 A CN 113239469A
Authority
CN
China
Prior art keywords
sample
structural parameter
network model
structural
vehicle body
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110662295.9A
Other languages
Chinese (zh)
Other versions
CN113239469B (en
Inventor
程然
林剑清
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southwest University of Science and Technology
Original Assignee
Southwest University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Southwest University of Science and Technology filed Critical Southwest University of Science and Technology
Priority to CN202110662295.9A priority Critical patent/CN113239469B/en
Publication of CN113239469A publication Critical patent/CN113239469A/en
Application granted granted Critical
Publication of CN113239469B publication Critical patent/CN113239469B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/17—Mechanical parametric or variational design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Abstract

The embodiment of the invention discloses a method, a device, equipment and a storage medium for optimizing the structure of a vehicle body part, wherein the method comprises the following steps: acquiring a generated countermeasure network model according to a structural parameter sample set of a vehicle body part, further acquiring a corresponding target generated countermeasure sample, and adding the target generated countermeasure sample into the structural parameter sample set; if the structural parameter sample set is judged not to accord with the preset iteration convergence condition, continuing to obtain the generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition; and according to at least one performance evaluation item, obtaining an optimal sample from the structural parameter sample set, and taking the optimal sample as a structural parameter of the vehicle body part, so that the structural optimization of the vehicle body part is realized, and meanwhile, a countermeasure sample is generated by generating a countermeasure network model to obtain a target generation countermeasure sample, so that the performance test cost of the vehicle body part is reduced, and the structural optimization efficiency of the vehicle body part is further improved.

Description

Structure optimization method, device, equipment and storage medium for vehicle body parts
Technical Field
The embodiment of the invention relates to the technical field of automobile manufacturing, in particular to a method, a device, equipment and a storage medium for optimizing the structure of a vehicle body part.
Background
Many problems such as environmental pollution and hidden traffic troubles caused by the continuous increase of the automobile holding capacity are always the hot points concerned by the development of the automobile industry and the social sustainable development at present. Most of the energy consumption of the automobile is consumed by the weight of the automobile, and the energy consumption of the automobile and the emission of greenhouse gases can be effectively reduced by reducing the weight of the automobile body. Therefore, the design of automobile light weight becomes a research focus in the field of automobile structure design at present.
The existing automobile lightweight design method mainly comprises an optimization design method based on a proxy model. Firstly, determining design variables and a design domain according to the design requirements of the vehicle body parts, then sampling in the design domain and carrying out performance test (evaluation of a design scheme) on samples, constructing an agent model according to the samples subjected to the performance test, and continuously updating the agent model by combining an optimization algorithm with a point adding criterion until the optimal vehicle body part parameters are obtained. However, this method requires a large number of iterations and a large number of performance tests, which is time-consuming and costly, and the proxy model often requires manual parameter setting, and the degree of automation of the design process is low.
Disclosure of Invention
The embodiment of the invention provides a method, a device, equipment and a storage medium for optimizing the structure of a vehicle body part, so as to realize the structural optimization of the vehicle body part.
In a first aspect, an embodiment of the present invention provides a method for optimizing a structure of a vehicle body component, including:
acquiring a generated countermeasure network model according to a structural parameter sample set of a vehicle body part, and acquiring at least one generated countermeasure sample according to the generated countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
judging whether the structural parameter sample set meets a preset iteration convergence condition or not;
if the structural parameter sample set does not accord with the preset iteration convergence condition, continuing to acquire a generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition;
and acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
In a second aspect, an embodiment of the present invention provides a device for optimizing a structure of a vehicle body component, including:
the system comprises a generation countermeasure sample acquisition module, a generation countermeasure network acquisition module and a generation countermeasure sample acquisition module, wherein the generation countermeasure sample acquisition module is used for acquiring a generation countermeasure network model according to a structural parameter sample set of a vehicle body part and acquiring at least one generation countermeasure sample according to the generation countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
a structural parameter sample set obtaining module, configured to obtain a target generation countermeasure sample from the at least one generation countermeasure sample according to a performance evaluation rule, and add the target generation countermeasure sample to the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
the condition judgment module is used for judging whether the structural parameter sample set meets a preset iteration convergence condition or not;
the model training module is used for continuously acquiring a generated confrontation network model according to the structural parameter sample set if the structural parameter sample set does not accord with a preset iteration convergence condition until the structural parameter sample set accords with the preset iteration convergence condition;
and the optimal sample acquisition module is used for acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
In a third aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:
one or more processors;
storage means for storing one or more programs;
when the one or more programs are executed by the one or more processors, the one or more processors implement the method for optimizing a structure of a vehicle body part according to any embodiment of the present invention.
In a fourth aspect, the embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the computer program implements the method for optimizing the structure of the vehicle body component according to any embodiment of the present invention.
According to the technical scheme provided by the embodiment of the invention, after a generated countermeasure network model is obtained according to a structural parameter sample set of a vehicle body part, at least one generated countermeasure sample is obtained according to the generated countermeasure network model, and then a corresponding target generated countermeasure sample is obtained, and the target generated countermeasure sample is added into the structural parameter sample set; if the structural parameter sample set is judged not to accord with the preset iteration convergence condition, continuing to obtain the generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition; and according to at least one performance evaluation item, obtaining an optimal sample from the structural parameter sample set, and taking the optimal sample as the structural parameter of the vehicle body part, thereby realizing the structural optimization of the vehicle body part, and simultaneously obtaining a target generation countermeasure sample by generating an countermeasure network model, and obtaining the structural parameter of the vehicle body part according with the performance evaluation rule only by a few times of performance tests, thereby improving the iterative convergence speed of the structural parameter sample set, reducing the performance test cost of the vehicle body part, and further improving the structural optimization efficiency of the vehicle body part.
Drawings
Fig. 1 is a flowchart of a method for optimizing a structure of a vehicle body component according to an embodiment of the present invention;
fig. 2A is a flowchart of a method for optimizing a structure of a vehicle body component according to a second embodiment of the present invention;
FIG. 2B is a schematic diagram of obtaining a model for generating a countermeasure network structure according to a second embodiment of the present invention;
fig. 3 is a structural block diagram of a structural optimization device for vehicle body parts according to a third embodiment of the present invention;
fig. 4 is a block diagram of an electronic device according to a fourth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a method for optimizing a structure of a vehicle body component according to an embodiment of the present invention, where the embodiment is applicable to optimizing a structure of a vehicle body component when designing a vehicle component, and the method can be executed by a device for optimizing a structure of a vehicle body component according to an embodiment of the present invention, where the device can be implemented by software and/or hardware and is integrated on an electronic device, and the method specifically includes the following steps:
s110, acquiring a generated countermeasure network model according to a structural parameter sample set of the vehicle body part, and acquiring at least one generated countermeasure sample according to the generated countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples.
The structural parameter sample describes structural parameters of a vehicle body part (such as a bumper, a front side rail and a roof side rail), and comprises a parameter type influencing the structure of the part and specific parameter values under the parameter type, wherein the parameter type corresponds to the parameter values one by one; for example, the vehicle body component is a roof rail, and the types of parameters affecting the roof rail structure include length, width, and thickness, which correspond to parameter values of 2 meters, 0.1 meters, and 0.1 meters, respectively, so that one sample of the structural parameters for the roof rail can be expressed as: the length is 2 meters, the width is 0.1 meter, and the thickness is 0.1 meter; the structural parameter sample set is a set formed by one or more structural parameter samples under the same vehicle body part, and the parameter values corresponding to the parameter types of the structural parameter samples can be assigned in advance by vehicle body designers according to experience or actual values of designed vehicle types.
Automobiles include various body parts, such as bumpers, front rails, and roof rails in the above-described embodiments; different vehicle body parts have different parameter types influencing the structure, for example, the parameter types corresponding to the bumper comprise length, width, thickness and bending degree; the corresponding parameter types of the roof carling comprise length, width and thickness; the method comprises the steps that vehicle body parts correspond to a structural parameter sample set one by one, and after a target vehicle body part (such as a roof rail) with parameters to be obtained at present is determined, the structural parameter sample set corresponding to the target vehicle body part is obtained; particularly, the number of the structural parameter samples in each structural parameter sample set may be preset as needed, the number of the structural parameter samples in the structural parameter sample sets corresponding to different vehicle body components is different, and the more complex the structure of the vehicle body component is (for example, the more parameter types corresponding to the vehicle body component are), the more the number of the structural parameter samples in the structural parameter sample set of the vehicle body component needs to be, so as to expand the selectable range of the actual parameter of the vehicle body component; meanwhile, according to different body parts, a structural parameter sample set comprising a corresponding number of structural parameter samples is obtained, and the acquisition of a confrontation network model can be generated more accurately.
Generating a countermeasure network (GAN) model, which is a deep learning model constructed based on a neural network, and comprises a generator and a discriminator, wherein the generator is used for generating an intermediate sample according to input random data which obeys preset probability distribution, and the discriminator is used for judging the intermediate sample generated by the generator and outputting a judgment result of the intermediate sample; through the combination of the generator and the discriminator, a more real generated confrontation sample can be obtained, and the accuracy of the obtained generated confrontation sample is improved. Optionally, in an embodiment of the present invention, the generator includes three hidden network layers, and the discriminator includes two hidden network layers.
The generated confrontation network model after training is a generated confrontation network model obtained by extracting a structure parameter sample from a structure parameter sample set as a training sample and training the initially generated confrontation network model by adopting the obtained training sample; wherein, the initially generated confrontation network model is constructed based on a neural network; and inputting the trained generated confrontation network model into random data which obeys preset probability distribution, and outputting the generated confrontation network model into generated confrontation samples which comprise parameter types consistent with the structural parameter samples and parameter values corresponding to the parameter types. Through the trained generated countermeasure network model, generated countermeasure samples corresponding to the current automobile body part can be generated, meanwhile, corresponding target generated countermeasure samples are selected from the generated countermeasure samples and added into the structural parameter sample set, and the number of samples in the structural parameter sample set is increased.
Generating a countermeasure sample, namely a structural parameter sample obtained by training a generated countermeasure network model; the generated countermeasure sample corresponding to the current vehicle body part is obtained through the generated countermeasure network model, so that the automatic obtaining of the generated countermeasure sample corresponding to the vehicle body part is realized, and the obtaining efficiency of the generated countermeasure sample is improved. Particularly, in the embodiment of the invention, a layer of network is added on the basis of the existing discriminator to predict the performance test result of the intermediate sample generated by the generator, so that the discriminator can predict the performance test result of the intermediate sample while judging the intermediate sample generated by the generator; by adding the performance test result prediction network layer, the prediction performance test result corresponding to each generated countermeasure sample can be obtained through the generated countermeasure network model, so that the times of performing performance test on the generated countermeasure samples can be reduced, the time cost for obtaining the structural parameters of the vehicle body parts is reduced, and the efficiency for obtaining the structural parameters of the vehicle body parts is improved.
Optionally, in the embodiment of the present invention, before obtaining and generating the countermeasure network model according to the structural parameter sample set of the vehicle body component, the method may further include: acquiring at least one parameter type and at least one candidate parameter value set of the vehicle body part, and assigning values to the at least one parameter type according to the at least one candidate parameter value set to acquire an assigned structural parameter sample set; and the parameter types correspond to the candidate parameter value sets one by one. Specifically, according to a vehicle body part to be optimized, a parameter type corresponding to the vehicle body part and an alternative parameter value set corresponding to the parameter type are obtained; the number of the parameter types may be one or multiple, and each parameter type corresponds to one candidate parameter value set.
The set of alternative parameter values, i.e. the set of selectable parameter values corresponding to the parameter type, may be a series of parameter values, or may be selectable range values of the parameter values. After the parameter type and the corresponding alternative parameter value set are determined, parameter value sampling is carried out in the alternative parameter value set, and the parameter value obtained by sampling is used for carrying out assignment on the corresponding parameter type so as to obtain an assigned structural parameter sample; the structural parameter sample set is a sample set composed of a plurality of assigned structural parameter samples. In particular, the parameter value sampling method comprises a Latin hypercube sampling method; latin Hypercube Sampling (LHS) is a layered random sampling method, efficient sampling can be carried out from a variable distribution interval, parameter value sampling is carried out in an alternative parameter value set by adopting a Latin hypercube sampling method, the obtained parameter values can be guaranteed to be distributed more uniformly, and the accuracy of structural parameter samples is further guaranteed.
After the assigned structural parameter sample set is obtained, performing performance test on each structural parameter sample in the structural parameter sample set to obtain a performance test result (for example, mass, energy absorption, specific energy absorption and peak force) corresponding to each structural parameter sample. If each structural parameter sample only corresponds to one performance index, directly taking the current performance index as a performance evaluation item, and sequencing all the structural parameter samples to obtain a sequenced structural parameter sample set; and if each structural parameter sample corresponds to a plurality of performance indexes, taking the performance indexes as performance evaluation items, and sequencing the structural parameter samples to obtain a sequenced structural parameter sample set. By obtaining the performance indexes corresponding to the structural parameter samples and sequencing the structural parameter samples according to the performance indexes, the performance quality condition corresponding to each structural parameter sample can be obtained, and further the classification of the structural parameter samples and the obtaining of the optimal sample corresponding to the optimal performance index are conveniently realized.
S120, according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set; wherein the performance evaluation rule includes at least one performance evaluation item.
The performance evaluation rule is an evaluation rule adopted for evaluating the generated countermeasure sample, and comprises performance evaluation according to a certain performance evaluation item and performance evaluation according to a plurality of performance evaluation items; wherein, the performance evaluation item is a performance index used for performance evaluation; the method comprises the steps of sequencing the generated countermeasure samples according to a performance evaluation rule to obtain one or more generated countermeasure samples with optimal performance exceeding a preset performance threshold value, and using the one or more generated countermeasure samples as target generated countermeasure samples, so that the target generated countermeasure samples are obtained.
Sequencing the obtained generated countermeasure samples according to the prediction performance test result corresponding to each generated countermeasure sample, and obtaining one or a preset number of generated countermeasure samples with optimal performance as target generated countermeasure samples according to the sequencing result; and performing performance test on the obtained target generation countermeasure samples, obtaining the performance test result corresponding to each target generation countermeasure sample, and finally adding the obtained target generation samples into the structural parameter sample set. One or a preset number of samples with optimal performance are obtained from the generated countermeasure samples and used as targets to generate the countermeasure samples for performance testing, so that the times of performance testing can be effectively reduced, the cost of performance testing of the vehicle body parts is reduced, and the structure optimization efficiency of the vehicle body parts is further improved.
Optionally, in this embodiment of the present invention, if the performance evaluation rule includes a plurality of performance evaluation items, the obtaining, according to the performance evaluation rule, a target generated countermeasure sample from the at least one generated countermeasure sample may include: and according to each performance evaluation item of the performance evaluation rule, performing non-dominant sorting on the at least one generated confrontation sample, performing re-sorting through a crowding distance in each dominant grade, and acquiring a target generated confrontation sample in the at least one generated confrontation sample. Specifically, when the performance evaluation rule includes a plurality of performance evaluation items, firstly, performing non-domination sorting on each generated countermeasure sample to obtain a domination level corresponding to each generated countermeasure sample; then, in each dominance level, that is, in the generated countermeasure samples at the same dominance level, sorting is performed again by the congestion distance, thereby realizing sorting of all the generated countermeasure samples.
The non-dominant ranking refers to dividing the generated confrontation samples into a dominant layer and a non-dominant layer according to a plurality of performance evaluation items, and dividing the non-dominant layer samples again until the samples can not be distinguished any more so as to obtain a dominant grade corresponding to each generated confrontation sample; the crowding distance is an average distance of corresponding parameter values of two generated confrontation samples on two adjacent sides of a certain generated confrontation sample, and the larger the crowding distance is, the more sparse the distribution of the currently generated confrontation samples is, and the higher the priority is. By the method, the antagonistic samples are accurately sequenced when a plurality of performance evaluation items exist, and the acquisition of the target antagonistic samples is further realized.
And S130, judging whether the structural parameter sample set meets a preset iteration convergence condition.
After the structural parameter sample set is obtained, at the moment, the structural parameter sample set comprises structural parameter samples and target generation countermeasure samples, and whether the current structural parameter sample set meets a preset iteration convergence condition is judged; the preset iteration convergence condition is a preset termination condition used for judging whether the acquired structural parameter sample set meets the design requirements of the vehicle body part, for example, the preset iteration convergence condition is a preset iteration execution time; if the current structural parameter sample set meets the preset iteration convergence condition, a corresponding optimal sample can be obtained according to the current structural parameter sample set, and if the current structural parameter sample set does not meet the preset iteration convergence condition, the structural parameter sample set is required to be adopted to generate a countermeasure network model for new training until the obtained structural parameter sample set meets the preset iteration convergence condition. By setting the preset iteration convergence condition, the acquired structural parameter sample set can meet the structural optimization requirements of the vehicle body parts.
Optionally, in this embodiment of the present invention, the determining whether the structural parameter sample set meets a preset iteration convergence condition may include: and judging whether the number of samples in the structural parameter sample set is greater than or equal to a preset number threshold, and/or judging whether the target generation countermeasure samples in the structural parameter sample set meet a preset performance evaluation threshold. Specifically, when the preset iteration convergence condition is a preset iteration number threshold, the number of target generation countermeasure samples obtained by each iteration is equal, so that the currently executed iteration number can be obtained according to the number of samples in the structural parameter sample set; therefore, by setting a preset iteration number threshold, when the number of samples in the structural parameter sample set is greater than or equal to the preset number threshold, the structural parameter sample set meeting the requirement can be obtained by representing that enough iteration times are executed currently.
And when the preset iteration convergence condition is a preset performance evaluation threshold, judging whether a sample with a corresponding performance index exceeding the preset performance evaluation threshold exists in the current parameter set, if so, indicating that an optimal sample meeting the task requirement can be obtained according to the current structural parameter sample set and used as the structural parameter of the vehicle body part. Particularly, any one of the two convergence conditions can be selected as a preset iteration convergence condition, and the structural parameter sample set only needs to meet the corresponding preset iteration convergence condition; the two convergence conditions can also be simultaneously selected as the preset iteration convergence conditions, and the structural parameter sample set can be determined to meet the preset iteration convergence conditions only if the structural parameter sample set needs to meet the two convergence conditions simultaneously, so that the flexibility and diversity of judging the structural parameter sample set are increased.
And S140, if the structural parameter sample set does not accord with the preset iterative convergence condition, continuing to acquire the generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iterative convergence condition.
Specifically, when the current structural parameter sample set is judged not to meet the preset iteration convergence condition, the training samples are extracted again from the current structural parameter sample set, and the preset generated confrontation network model is trained again to obtain a new generated confrontation network model; and acquiring a new target generation countermeasure sample according to the new generation countermeasure network model, adding the new target generation countermeasure sample into the structural parameter sample set, and repeating the steps until the acquired structural parameter sample set meets the preset iteration convergence condition.
Optionally, in the embodiment of the present invention, after determining whether the structural parameter sample set meets a preset iteration convergence condition, the method may further include: and if the structural parameter sample set meets a preset iteration convergence condition, acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part. Specifically, when the structural parameter sample set meets a preset iteration convergence condition, sequencing samples in the structural parameter sample set according to one or more performance evaluation items; the specific sorting method is consistent with the above method for obtaining the target to generate the confrontation sample according to the performance evaluation rule, and is not described herein again. After the sorted structural parameter sample set is obtained, selecting a sample with the optimal performance as an optimal sample, and taking the structural parameter type corresponding to the optimal sample and the matched parameter value as structural parameters of the vehicle body part; in particular, the optimal samples may be structural parameter samples, and countermeasure samples may be generated for the target.
S150, obtaining an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
After iteration is carried out for a plurality of times until the obtained structural parameter sample set meets a preset iteration convergence condition, samples in the current structural parameter sample set are sequenced according to one or more performance evaluation items to obtain samples with optimal performance as optimal samples, and the structural parameter types contained in the optimal samples and the parameter values corresponding to the structural parameter types are used as the structural parameters of the vehicle body parts, so that the structural optimization of the vehicle body parts is realized.
According to the technical scheme provided by the embodiment of the invention, after a generated countermeasure network model is obtained according to a structural parameter sample set of a vehicle body part, at least one generated countermeasure sample is obtained according to the generated countermeasure network model, and then a corresponding target generated countermeasure sample is obtained, and the target generated countermeasure sample is added into the structural parameter sample set; if the structural parameter sample set is judged not to accord with the preset iteration convergence condition, continuing to obtain the generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition; and according to at least one performance evaluation item, obtaining an optimal sample from the structural parameter sample set, and taking the optimal sample as the structural parameter of the vehicle body part, thereby realizing the structural optimization of the vehicle body part, and simultaneously obtaining a target generation countermeasure sample by generating an countermeasure network model, and obtaining the structural parameter of the vehicle body part according with the performance evaluation rule only by a few times of performance tests, thereby improving the iterative convergence speed of the structural parameter sample set, reducing the performance test cost of the vehicle body part, and further improving the structural optimization efficiency of the vehicle body part.
Example two
Fig. 2A is a flowchart of a method for optimizing a structure of a vehicle body component according to a second embodiment of the present invention, which is embodied on the basis of the second embodiment, in the present embodiment, a countermeasure network model is obtained and generated according to a structure parameter sample set, so as to optimize the structure of the vehicle body component, where the method specifically includes:
s210, performing performance test on each structural parameter sample in the structural parameter sample set, classifying each structural parameter sample according to the performance evaluation rule to obtain a positive structural parameter sample and a negative structural parameter sample, and executing S220.
The performance test is to perform a simulation test (for example, a collision test) on the vehicle body parts corresponding to the structural parameter samples to obtain at least one performance index corresponding to each structural parameter sample; in particular, the performance test is performed once, and the time taken is long; the performance test method comprises a finite element analysis method, wherein the finite element analysis method is used for simulating a real system (a vehicle system) by using a mathematical method so as to realize simulation analysis of the real system; through performance test on each structural parameter sample, the performance indexes (such as mass, hardness and peak force) corresponding to each structural parameter sample are obtained.
After the performance parameters corresponding to the structural parameter samples are obtained, sequencing all the structural parameter samples according to a performance evaluation rule; optionally, in the embodiment of the present invention, if the sorted structural parameter sample set includes an even number of structural parameter samples, the first half of the structural parameter samples are selected as positive structural parameter samples, and the second half of the structural parameter samples are selected as negative structural parameter samples; if the sorted structural parameter sample set comprises odd structural parameter samples, aiming at the sorted structural parameter samples, a preset number of structural parameter samples can be selected as positive structural parameter samples at the front part, and a preset number of structural parameter samples are selected as negative structural parameter samples at the rear part; therefore, flexible classification of the structural parameter samples in the structural parameter sample set is realized, and the positive and negative structural parameter samples with equal number are obtained.
Optionally, in this embodiment of the present invention, if the performance evaluation rule includes a plurality of performance evaluation items, the classifying, according to the performance evaluation rule, each of the structural parameter samples to obtain a positive structural parameter sample and a negative structural parameter sample may include: and according to each performance evaluation item of the performance evaluation rule, performing non-dominant sorting on each structural parameter sample, and performing re-sorting through crowding distance in each dominant grade to classify each structural parameter sample to obtain a positive structural parameter sample and a negative structural parameter sample. The non-dominated sorting and the congestion distance have been described in detail in the first embodiment, and are not described herein again. After the structural parameter samples are sorted, as shown in fig. 2B, the same number of positive structural parameter samples and negative structural parameter samples are obtained in the structural parameter sample set; in the figure, the sample set represents a structural parameter sample set, the good samples represent positive structural parameter samples, and the bad samples represent negative structural parameter samples.
S220, training the initially generated confrontation network model according to the positive structure parameter sample and the negative structure parameter sample to obtain an intermediate generated confrontation network model, and executing S230.
Specifically, an initially generated confrontation network model is constructed according to a neural network, and the initially generated confrontation network model is trained according to a positive structure parameter sample and a negative structure parameter sample; specifically, as shown in fig. 2B, noise (random data) that obeys the probability distribution of positive structure parameter samples is used as an input of a generator that generates a countermeasure network model, and countermeasure samples are generated through the output of the generator; and inputting the positive structural parameter sample, the negative structural parameter sample and the intermediate generation countermeasure sample into a discriminator, realizing positive and negative judgment on the intermediate generation countermeasure sample through the discriminator, and simultaneously outputting a prediction performance test result corresponding to the intermediate generation countermeasure sample.
And S230, outputting an intermediate generation countermeasure sample through the generator of the intermediate generation countermeasure network model, and executing S240.
S240, judging whether the intermediate generation countermeasure sample is a positive structure parameter sample or not through the discriminator of the intermediate generation countermeasure network model.
Specifically, when the discriminator judges that the intermediate generation confrontation samples output by the generator are all positive structure parameter samples, the training of the generation confrontation network model is completed; therefore, the initially generated confrontation network model needs to be iteratively trained for multiple times until the trained generated confrontation network model is obtained. Generating an intermediate confrontation network model, namely generating the confrontation network model without training, namely acquiring the generated confrontation network model after training the initially generated confrontation network model for a certain number of times; by judging whether the intermediate generation countermeasure sample corresponding to the intermediate generation countermeasure network model is a positive structure parameter sample, the judgment of whether the training of the intermediate generation countermeasure network model is completed is realized. For training to generate the confrontation network model, an overall objective function for generating the confrontation network model is obtained based on the following formula:
Figure BDA0003115841340000151
where x represents the input positive and negative structural parameter samples, pdataRepresenting the probability distribution, x-p, obeyed by the positive and negative structural parameter samplesdataRepresenting sample data x obeying a probability distribution pdataZ represents random data input by the generator, pz(z) represents the probability distribution of samples of the positive structural parameter, z to pz(z) representing the probability distribution of the random data obeying positive structural parameter samples, G representing the generator, G (z) representing the output of the generator for input z,
Figure BDA0003115841340000152
denotes minimizing the generator output, D denotes the arbiter, D (x) denotes the arbiter output for input x,
Figure BDA0003115841340000153
it is meant that the output of the discriminator is maximized,
Figure BDA0003115841340000154
representing a mathematical expectation, arg represents having generator G obtain a minimum while discriminator D obtains a maximumThe corresponding parameter value.
Specifically, firstly, random data which obey the probability sample distribution of positive structure parameter samples is used as the input of a generator, and a group of intermediate generation countermeasure samples are generated through the generator; fixing the parameters of the generator G, generating a confrontation sample training discriminator D by adopting a positive structure parameter sample, a negative structure parameter sample and the middle generated by the generator, and adjusting the parameters of the discriminator to maximize the overall objective function L; then fixing parameters of the discriminator D, training a generator G, and adjusting the parameters of the generator to minimize the mathematical expectation of log (1-D (G (z)); and repeating the process until all the intermediate generation confrontation samples generated by the decision generator G of the discriminator D are positive structure parameter samples, and stopping training. Iterative training is carried out on the initially generated confrontation network model according to the positive structure parameter sample and the negative structure parameter sample, and whether the intermediate generated confrontation sample is the positive structure parameter sample or not is judged so as to determine whether training is completed or not, and the generation of the confrontation network model after training is completed is achieved.
Optionally, in this embodiment of the present invention, the determining, by the identifier of the intermediate generation countermeasure network model, whether the intermediate generation countermeasure sample is a positive structure parameter sample may include: and if the classifier of the intermediate generation countermeasure network model judges the classification result of the intermediate generation countermeasure sample as a positive structure parameter sample and judges that the prediction performance test result of the intermediate generation countermeasure sample conforms to a preset test threshold, judging the intermediate generation countermeasure sample as a positive structure parameter sample. Specifically, as shown in fig. 2B, the discriminator outputs the result of the prediction performance test for the currently intermediately generated countermeasure sample while outputting the result of the determination for the intermediately generated countermeasure sample; therefore, when the discriminator judges the intermediate generation countermeasure sample, on the basis of judging whether the intermediate generation countermeasure sample is a positive structure parameter sample, judging whether the prediction performance test result of the intermediate generation countermeasure sample meets a preset test threshold value; only when the two judgment conditions are met, the training of the current initially generated confrontation network model is determined to be completed, and the accuracy of the acquired confrontation network model is improved.
If the intermediate generation countermeasure network model discriminator determines that the intermediate generation countermeasure sample is a positive structure parameter sample, then S250 is executed; and if the discriminator of the intermediate generation countermeasure network model judges that the intermediate generation countermeasure sample is not the positive structure parameter sample, returning to execute S220.
Specifically, if the discriminator of the intermediate generation confrontation network model judges that the acquired confrontation samples generated in the middle within a certain time or in a certain number of middle are positive structural parameter samples, it is possible to confirm that the training of the intermediate generation confrontation network model is completed, and use the current intermediate generation confrontation network model as the generated confrontation network model after the training is completed. And if the discriminator of the intermediate generation countermeasure network model judges that the intermediate generation countermeasure sample is not the positive structure parameter sample, training the intermediate generation countermeasure network model continuously according to the positive structure parameter sample and the negative structure parameter sample until the intermediate generation countermeasure sample output by the generator of the intermediate generation countermeasure network model is judged as the positive structure parameter sample by the discriminator of the intermediate generation countermeasure network model.
And S250, taking the intermediate generation confrontation network model as a training completed generation confrontation network model, and executing S260.
S260, obtaining at least one generated countermeasure sample according to the generated countermeasure network model; wherein the structural parameter sample set comprises a plurality of structural parameter samples, S270 is executed.
S270, according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set, wherein the performance evaluation rule comprises at least one performance evaluation item; s280 is performed.
And S280, judging whether the structural parameter sample set meets a preset iteration convergence condition.
If the structural parameter sample set does not meet the preset iteration convergence condition, returning to execute S260; specifically, the generated confrontation network model is continuously obtained according to the structural parameter sample set until the structural parameter sample set meets a preset iteration convergence condition. If the structural parameter sample set meets the predetermined iteration convergence condition, S290 is performed.
S290, obtaining an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
According to the technical scheme provided by the embodiment of the invention, the structural parameter samples in the structural parameter sample set are classified to obtain positive structural parameter samples and negative structural parameter samples, and the initially generated confrontation network model is trained by adopting the positive structural parameter samples and the negative structural parameter samples to obtain the trained generated confrontation network model; the acquisition of the accurately generated confrontation network model is realized, and the accuracy of the acquired confrontation sample generation is improved; meanwhile, the countermeasure sample generated corresponding to the target is obtained according to the generated countermeasure network model after training, the optimal sample is obtained according to the target generation countermeasure sample and the structural parameter sample set, the optimal sample is used as the structural parameter of the vehicle body part, structural optimization of the vehicle body part is achieved, meanwhile, the target generation countermeasure sample is obtained through the generated countermeasure network model, the structural parameter of the vehicle body part meeting the performance evaluation rule can be obtained only through performance tests of fewer times, the iterative convergence speed of the structural parameter sample set is improved, the performance test cost of the vehicle body part is reduced, and the structural optimization efficiency of the vehicle body part is further improved.
EXAMPLE III
Fig. 3 is a structural block diagram of a structural optimization device for vehicle body parts according to a third embodiment of the present invention, where the device specifically includes: a generation confrontation sample acquisition module 310, a structural parameter sample set acquisition module 320, a condition judgment module 330, a model training module 340 and an optimal sample acquisition module 350;
a generation countermeasure sample acquisition module 310, configured to acquire a generation countermeasure network model according to a structural parameter sample set of a vehicle body component, and acquire at least one generation countermeasure sample according to the generation countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
a structural parameter sample set obtaining module 320, configured to obtain a target generation countermeasure sample from the at least one generation countermeasure sample according to a performance evaluation rule, and add the target generation countermeasure sample to the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
a condition determining module 330, configured to determine whether the structural parameter sample set meets a preset iteration convergence condition;
the model training module 340 is configured to, if the structural parameter sample set does not meet a preset iteration convergence condition, continue to obtain a generated confrontation network model according to the structural parameter sample set until the structural parameter sample set meets the preset iteration convergence condition;
and an optimal sample obtaining module 350, configured to obtain an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and use the optimal sample as a structural parameter of the vehicle body component.
According to the technical scheme provided by the embodiment of the invention, after a generated countermeasure network model is obtained according to a structural parameter sample set of a vehicle body part, at least one generated countermeasure sample is obtained according to the generated countermeasure network model, and then a corresponding target generated countermeasure sample is obtained, and the target generated countermeasure sample is added into the structural parameter sample set; if the structural parameter sample set is judged not to accord with the preset iteration convergence condition, continuing to obtain the generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition; and according to at least one performance evaluation item, obtaining an optimal sample from the structural parameter sample set, and taking the optimal sample as the structural parameter of the vehicle body part, thereby realizing the structural optimization of the vehicle body part, and simultaneously obtaining a target generation countermeasure sample by generating an countermeasure network model, and obtaining the structural parameter of the vehicle body part according with the performance evaluation rule only by a few times of performance tests, thereby improving the iterative convergence speed of the structural parameter sample set, reducing the performance test cost of the vehicle body part, and further improving the structural optimization efficiency of the vehicle body part.
Optionally, on the basis of the foregoing technical solution, the optimal sample obtaining module 350 is further configured to obtain an optimal sample from the structural parameter sample set according to the at least one performance evaluation item if the structural parameter sample set meets a preset iterative convergence condition, and use the optimal sample as the structural parameter of the vehicle body component.
Optionally, on the basis of the foregoing technical solution, the condition determining module 330 is specifically configured to determine whether the number of samples in the structural parameter sample set is greater than or equal to a preset number threshold, and/or determine whether the target generation countermeasure sample in the structural parameter sample set meets a preset performance evaluation threshold.
Optionally, on the basis of the above technical solution, the structure optimization device for vehicle body components further includes:
the system comprises a set acquisition module, a parameter model generation module and a parameter model generation module, wherein the set acquisition module is used for acquiring at least one parameter type and at least one candidate parameter value set of the vehicle body parts, and assigning values to the at least one parameter type according to the at least one candidate parameter value set so as to acquire an assigned structural parameter sample set; and the parameter types correspond to the candidate parameter value sets one by one.
Optionally, on the basis of the foregoing technical solution, the generation countermeasure sample obtaining module 310 includes:
the sample classification unit is used for carrying out performance test on each structural parameter sample in the structural parameter sample set and classifying each structural parameter sample according to the performance evaluation rule so as to obtain a positive structural parameter sample and a negative structural parameter sample;
the model training unit is used for training the initially generated confrontation network model according to the positive structure parameter sample and the negative structure parameter sample so as to obtain an intermediate generated confrontation network model;
the sample judging unit is used for outputting an intermediate generation countermeasure sample through the generator of the intermediate generation countermeasure network model and judging whether the intermediate generation countermeasure sample is a positive structure parameter sample or not through the discriminator of the intermediate generation countermeasure network model;
and the model acquisition unit is used for taking the intermediate generation confrontation network model as the generation confrontation network model after training if the discriminator of the intermediate generation confrontation network model judges that the intermediate generation confrontation sample is the positive structure parameter sample.
Optionally, on the basis of the above technical solution, the model training unit is further configured to, if the discriminator of the intermediate generation confrontation network model determines that the intermediate generation confrontation sample is not the positive structure parameter sample, continue training the intermediate generation confrontation network model according to the positive structure parameter sample and the negative structure parameter sample;
and the model acquisition unit is further used for taking the intermediate generation confrontation network model as the training-finished generation confrontation network model when the intermediate generation confrontation sample output by the generator of the intermediate generation confrontation network model is judged as a positive structure parameter sample by the discriminator of the intermediate generation confrontation network model.
Optionally, on the basis of the foregoing technical solution, the structure parameter sample set obtaining module 320 is specifically configured to, if the performance evaluation rule includes a plurality of performance evaluation items, perform non-dominant ranking on the at least one generated countermeasure sample according to each performance evaluation item of the performance evaluation rule, perform re-ranking in each dominant rank through a congestion distance, and obtain a target generated countermeasure sample in the at least one generated countermeasure sample.
Optionally, on the basis of the foregoing technical solution, the sample classification unit is specifically configured to, if the performance evaluation rule includes a plurality of performance evaluation items, perform non-dominated sorting on each structural parameter sample according to each performance evaluation item of the performance evaluation rule, and perform re-sorting through a congestion distance in each domination level, so as to classify each structural parameter sample, and obtain a positive structural parameter sample and a negative structural parameter sample.
Optionally, on the basis of the foregoing technical solution, the sample determining unit is specifically configured to determine that the intermediate generation countermeasure sample is a positive structure parameter sample if the classifier of the intermediate generation countermeasure network model determines that the classification result of the intermediate generation countermeasure sample is a positive structure parameter sample, and determines that the prediction performance test result of the intermediate generation countermeasure sample meets a preset test threshold.
The device can execute the structure optimization method of the vehicle body part provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details not described in detail in this embodiment, reference may be made to the method provided in any embodiment of the present invention.
Example four
Fig. 4 is a schematic structural diagram of an electronic device according to a fourth embodiment of the present invention. FIG. 4 illustrates a block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention. The electronic device 12 shown in fig. 4 is only an example and should not bring any limitation to the function and the scope of use of the embodiment of the present invention.
As shown in FIG. 4, electronic device 12 is embodied in the form of a general purpose computing device. The components of electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a memory 28, and a bus 18 that couples various system components including the memory 28 and the processing unit 16.
Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, such architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, micro-channel architecture (MAC) bus, enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.
Electronic device 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by electronic device 12 and includes both volatile and nonvolatile media, removable and non-removable media.
The memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)30 and/or cache memory 32. The electronic device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 4, and commonly referred to as a "hard drive"). Although not shown in FIG. 4, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.
A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 42 generally carry out the functions and/or methodologies of the described embodiments of the invention.
Electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), with one or more devices that enable a user to interact with electronic device 12, and/or with any devices (e.g., network card, modem, etc.) that enable electronic device 12 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 22. Also, the electronic device 12 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via the network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via the bus 18. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
The processing unit 16 executes various functional applications and data processing by executing programs stored in the memory 28, for example, to implement a method for optimizing the structure of a vehicle body component according to any of the embodiments of the present invention. Namely: acquiring a generated countermeasure network model according to a structural parameter sample set of a vehicle body part, and acquiring at least one generated countermeasure sample according to the generated countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples; according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item; judging whether the structural parameter sample set meets a preset iteration convergence condition or not; if the structural parameter sample set does not accord with the preset iteration convergence condition, continuing to acquire a generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition; and acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
EXAMPLE five
Fifth, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements a method for optimizing a structure of a vehicle body component according to any embodiment of the present invention; the method comprises the following steps:
acquiring a generated countermeasure network model according to a structural parameter sample set of a vehicle body part, and acquiring at least one generated countermeasure sample according to the generated countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
judging whether the structural parameter sample set meets a preset iteration convergence condition or not;
if the structural parameter sample set does not accord with the preset iteration convergence condition, continuing to acquire a generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition;
and acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims (10)

1. A method for optimizing the structure of a vehicle body part is characterized by comprising the following steps:
acquiring a generated countermeasure network model according to a structural parameter sample set of a vehicle body part, and acquiring at least one generated countermeasure sample according to the generated countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
according to a performance evaluation rule, obtaining a target generation countermeasure sample from the at least one generation countermeasure sample, and adding the target generation countermeasure sample into the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
judging whether the structural parameter sample set meets a preset iteration convergence condition or not;
if the structural parameter sample set does not accord with the preset iteration convergence condition, continuing to acquire a generated confrontation network model according to the structural parameter sample set until the structural parameter sample set accords with the preset iteration convergence condition;
and acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
2. The method of claim 1, after determining whether the set of structural parameter samples meets a predetermined iteration convergence condition, further comprising:
and if the structural parameter sample set meets a preset iteration convergence condition, acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
3. The method of claim 1, wherein the determining whether the set of structural parameter samples meets a predetermined iteration convergence condition comprises:
and judging whether the number of samples in the structural parameter sample set is greater than or equal to a preset number threshold, and/or judging whether the target generation countermeasure samples in the structural parameter sample set meet a preset performance evaluation threshold.
4. The method of claim 1, before obtaining and generating the countermeasure network model according to the structural parameter sample set of the vehicle body component, further comprising:
acquiring at least one parameter type and at least one candidate parameter value set of the vehicle body part, and assigning values to the at least one parameter type according to the at least one candidate parameter value set to acquire an assigned structural parameter sample set; and the parameter types correspond to the candidate parameter value sets one by one.
5. The method according to claim 1, wherein the obtaining a generative confrontation network model from a sample set of structural parameters of a vehicle body part comprises:
performing performance test on each structural parameter sample in the structural parameter sample set, and classifying each structural parameter sample according to the performance evaluation rule to obtain a positive structural parameter sample and a negative structural parameter sample;
training an initially generated confrontation network model according to the positive structure parameter sample and the negative structure parameter sample to obtain an intermediate generated confrontation network model;
outputting an intermediate generation countermeasure sample through the generator of the intermediate generation countermeasure network model, and judging whether the intermediate generation countermeasure sample is a positive structure parameter sample through a discriminator of the intermediate generation countermeasure network model;
if the discriminator of the intermediate generation confrontation network model judges that the intermediate generation confrontation sample is a positive structure parameter sample, taking the intermediate generation confrontation network model as a generation confrontation network model after training is finished;
if the intermediate generation countermeasure network model discriminator judges that the intermediate generation countermeasure sample is not a positive structure parameter sample, training the intermediate generation countermeasure network model is continued according to the positive structure parameter sample and the negative structure parameter sample until the intermediate generation countermeasure sample output by the intermediate generation countermeasure network model generator is judged to be a positive structure parameter sample, and when the intermediate generation countermeasure network model discriminator judges that the intermediate generation countermeasure network model is a training-completed generation countermeasure network model.
6. The method according to claim 5, wherein if the performance evaluation rule includes a plurality of performance evaluation items, the obtaining a target countermeasure sample among the at least one countermeasure sample according to the performance evaluation rule comprises:
according to each performance evaluation item of the performance evaluation rule, performing non-dominant sorting on the at least one generated confrontation sample, performing re-sorting through a crowding distance in each dominant grade, and acquiring a target generated confrontation sample in the at least one generated confrontation sample;
and/or classifying the structural parameter samples according to the performance evaluation rule to obtain positive structural parameter samples and negative structural parameter samples, including:
and according to each performance evaluation item of the performance evaluation rule, performing non-dominant sorting on each structural parameter sample, and performing re-sorting through crowding distance in each dominant grade to classify each structural parameter sample to obtain a positive structural parameter sample and a negative structural parameter sample.
7. The method of claim 5, wherein the determining whether the intermediate generated countermeasure sample is a positive structure parameter sample by the discriminator of the intermediate generated countermeasure network model comprises:
and if the classifier of the intermediate generation countermeasure network model judges the classification result of the intermediate generation countermeasure sample as a positive structure parameter sample and judges that the prediction performance test result of the intermediate generation countermeasure sample conforms to a preset test threshold, judging the intermediate generation countermeasure sample as a positive structure parameter sample.
8. A structural optimization device for vehicle body parts, comprising:
the system comprises a generation countermeasure sample acquisition module, a generation countermeasure network acquisition module and a generation countermeasure sample acquisition module, wherein the generation countermeasure sample acquisition module is used for acquiring a generation countermeasure network model according to a structural parameter sample set of a vehicle body part and acquiring at least one generation countermeasure sample according to the generation countermeasure network model; wherein the set of structural parameter samples comprises a plurality of structural parameter samples;
a structural parameter sample set obtaining module, configured to obtain a target generation countermeasure sample from the at least one generation countermeasure sample according to a performance evaluation rule, and add the target generation countermeasure sample to the structural parameter sample set; wherein the performance evaluation rule comprises at least one performance evaluation item;
the condition judgment module is used for judging whether the structural parameter sample set meets a preset iteration convergence condition or not;
the model training module is used for continuously acquiring a generated confrontation network model according to the structural parameter sample set if the structural parameter sample set does not accord with a preset iteration convergence condition until the structural parameter sample set accords with the preset iteration convergence condition;
and the optimal sample acquisition module is used for acquiring an optimal sample from the structural parameter sample set according to the at least one performance evaluation item, and taking the optimal sample as the structural parameter of the vehicle body part.
9. An electronic device, characterized in that the electronic device comprises:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method of structural optimization of a vehicle body part as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a method for structural optimization of a vehicle body part according to any one of claims 1 to 7.
CN202110662295.9A 2021-06-15 2021-06-15 Structure optimization method, device, equipment and storage medium for vehicle body parts Active CN113239469B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110662295.9A CN113239469B (en) 2021-06-15 2021-06-15 Structure optimization method, device, equipment and storage medium for vehicle body parts

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110662295.9A CN113239469B (en) 2021-06-15 2021-06-15 Structure optimization method, device, equipment and storage medium for vehicle body parts

Publications (2)

Publication Number Publication Date
CN113239469A true CN113239469A (en) 2021-08-10
CN113239469B CN113239469B (en) 2023-04-07

Family

ID=77139982

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110662295.9A Active CN113239469B (en) 2021-06-15 2021-06-15 Structure optimization method, device, equipment and storage medium for vehicle body parts

Country Status (1)

Country Link
CN (1) CN113239469B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109003678A (en) * 2018-06-12 2018-12-14 清华大学 A kind of generation method and system emulating text case history
CN109388565A (en) * 2018-09-27 2019-02-26 西安电子科技大学 Software system performance optimization method based on production confrontation network
CN109523493A (en) * 2017-09-18 2019-03-26 杭州海康威视数字技术股份有限公司 A kind of image generating method, device and electronic equipment
CN110542819A (en) * 2019-09-25 2019-12-06 贵州电网有限责任公司 transformer fault type diagnosis method based on semi-supervised DBNC
CN111476269A (en) * 2020-03-04 2020-07-31 中国平安人寿保险股份有限公司 Method, device, equipment and medium for constructing balanced sample set and identifying copied image
CN111598210A (en) * 2020-04-30 2020-08-28 浙江工业大学 Anti-attack defense method based on artificial immune algorithm
CN112465184A (en) * 2020-10-21 2021-03-09 广西大学 Cloud energy storage system control method of small-sample generation type counterstudy network

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109523493A (en) * 2017-09-18 2019-03-26 杭州海康威视数字技术股份有限公司 A kind of image generating method, device and electronic equipment
CN109003678A (en) * 2018-06-12 2018-12-14 清华大学 A kind of generation method and system emulating text case history
CN109388565A (en) * 2018-09-27 2019-02-26 西安电子科技大学 Software system performance optimization method based on production confrontation network
CN110542819A (en) * 2019-09-25 2019-12-06 贵州电网有限责任公司 transformer fault type diagnosis method based on semi-supervised DBNC
CN111476269A (en) * 2020-03-04 2020-07-31 中国平安人寿保险股份有限公司 Method, device, equipment and medium for constructing balanced sample set and identifying copied image
CN111598210A (en) * 2020-04-30 2020-08-28 浙江工业大学 Anti-attack defense method based on artificial immune algorithm
CN112465184A (en) * 2020-10-21 2021-03-09 广西大学 Cloud energy storage system control method of small-sample generation type counterstudy network

Also Published As

Publication number Publication date
CN113239469B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
US11100266B2 (en) Generating integrated circuit floorplans using neural networks
CN110399222B (en) GPU cluster deep learning task parallelization method and device and electronic equipment
CN113240155A (en) Method and device for predicting carbon emission and terminal
CN108446727B (en) Driving behavior decision method and system and electronic equipment
CN111191824A (en) Power battery capacity attenuation prediction method and system
CN112766402A (en) Algorithm selection method and device and electronic equipment
US10803218B1 (en) Processor-implemented systems using neural networks for simulating high quantile behaviors in physical systems
CN112883651B (en) System-level testability design multi-objective optimization method based on improved PBI method
CN116738920B (en) Chip 3D design method and system of three-dimensional geometric kernel
Amrit et al. Design strategies for multi-objective optimization of aerodynamic surfaces
Li et al. A comparative study of pre-screening strategies within a surrogate-assisted multi-objective algorithm framework for computationally expensive problems
CN110516164A (en) A kind of information recommendation method, device, equipment and storage medium
CN113239469B (en) Structure optimization method, device, equipment and storage medium for vehicle body parts
Zhao et al. Adaptive Swarm Intelligent Offloading Based on Digital Twin-assisted Prediction in VEC
Liu et al. An intelligent optimization method for highway route selection based on comprehensive weight and TOPSIS
CN111815209A (en) Data dimension reduction method and device applied to wind control model
CN115729829A (en) Automatic driving acceleration test method based on improved genetic algorithm
CN115166811A (en) Method for quickly constructing neutron spectrum in simulation working site
CN114064235A (en) Multitask teaching and learning optimization method, system and equipment
CN112241811A (en) Method for predicting hierarchical mixed performance of customized product in 'Internet +' environment
Yun et al. Metaheuristic-based inspection policy for a one-shot system with two types of units
JP2020123292A (en) Evaluation method of neural network, generation method of neural network, program, and evaluation system
CN115859768B (en) Method and device for predicting work piece finishing time of dynamic assembly workshop
CN116662853B (en) Method and system for automatically identifying analysis result of pollution source
CN115270584B (en) Lightweight method suitable for new energy electric vehicle battery bracket

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant